No. VI. ON WORKING IRON AND STEEL

Early Journal Content on JSTOR, Free to Anyone in the World

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in
the world by JSTOR.

Known as the Early Journal Content, this set of works include research articles, news, letters, and other
writings published in more than 200 of the oldest leading academic journals. The works date from the
mid-seventeenth to the early twentieth centuries.

We encourage people to read and share the Early Journal Content openly and to tell others that this
resource exists. People may post this content online or redistribute in any way for non-commercial
purposes.

Read more about Early Journal Content at http://about.istor.org/participate-istor/individuals/early-
journal-content .

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people
discover, use, and build upon a wide range of content through a powerful research and teaching
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please
contact support@jstor.org.

250 MECHANICS.

new, as well as the contrivance of hanging it on pivots, in
order to accommodate itself to the varying pressure of the
hand.

No. VI.

ON WORKING IRON AND STEEL.

The Thanks of the Society were presented to C. Var-
LEV, Esq. of], Charles Street, Clarendon Square, for
the following paper.

The Society having favourably received the description
of my late uncle's method of condensing brass, in which
I endeavoured to shew the conditions which are requisite
for the extreme and uniform condensation of metal pre-
viously sound ; and knowing of how much consequence
it would be to secure the soundness of anchors, iron ties,
girders, and other implements, to which the safety of
human life and property are so often trusted, I am en-
couraged to pursue the subject, and endeavour to shew
the means by which iron or steel may be rendered sound,
and preserved so while working, and even during welding.
But on a subject supposed to be so well known, and in
which I must repeat much that is known, for the sake of
connexion, it may be well to justify myself by shewing, in
the first place, how inadequate the ordinary practice is to
prevent unsoundness.

On Iron.

The welding of numerous layers of good iron together,
so as to form one bar, is considered to render it much
tougher and more trustworthy than if it had been wrought

MECHANICS. 261

out of one piece ; for it breaks joint with, and equally
mixes the strong and weak parts. But many stop here,
believing they have done all they can ; yet every welding
may be discovered, which proves that a considerable de-
gree of unsoundness accompanies this process.

If we notice the pump-handles about town which are
constantly exposed to the weather, some are found where
the continual wear from handling exceeds that of the
weather: these acquire nearly the best polish that soft
iron is capable of. But there are many in which the
wear from handling barely keeps pace with the decom-
position by weather ; in these, the last weldings which
united the different parts together to form the handle are
most distinctly marked, and the various layers of which
the mass has probably at different times been formed may
be seen, very much resembling the longitudinal grain of
wood. Here is sufficient proof of the want of homo-
geneity ; and, seeing it is a constant attendant on welded
iron, we are led to inquire what causes the necessity for
welding. The answer is, first, the acknowledgment that
iron is liable to invisible unsoundnesses, and that by weld-
ing they are likely to be distributed among sound parts
and thereby equalise the strength of the mass ; secondly,
that there is a certain thickness beyond which the most
governable hammers fail to produce the full effect of con-
densing the iron all through ; and this thickness being
far short of that required for anchors and other imple-
ments in which the utmost soundness and strength are
wanted, such can only be made up by welding. Within
that thickness there is no direct necessity for welding;
but beyond it we have no means of rendering the whole
mass sound. This necessity of sometimes having recourse
to welding renders it of consequence to know what there

252 MECHANICS.

is in the process of welding that makes the laminae
visible.

When newly reduced iron is hammered till it becomes
quite malleable and well closed or united, it may be con-
sidered homogeneal and sound, from its having no weld-
ings and no carbon, this latter being carried off by the
oxygen of the bar and beating. W hatever oxygen the
surface imbibes in after-heatings is beaten off in scales
while giving the required shape to the bar. It therefore
remains as sound and free from oxygen as before ; for oxide
of iron is not only brittle, but incapable of uniting by
welding with malleable iron.

But when two such masses or bars are heated for
welding, they become again coated with oxide ; and then,
when laid one upon the other and submitted to the action
of the hammer, the oxide is shut in, and union of the bars
can only take place in proportion as the two plates of
oxide are broken up and the subjacent malleable particles
are brought into mutual contact.

From this intermixture of oxide and metal arises a
spongy texture (more easily acted on by the air than
the pure solid metal), which distinguishes the welding
from the rest of the mass. If, therefore, several layers of
good iron were welded together and beaten very much
out without hammering the edges, on filing these square
and cleaning the four sides, a little weak acid would soon
distinguish the edges of the layers from their flat sides,
and they might probably be counted.

But in ordinary practice iron is never left so : it is
beaten in all directions, to give or keep the required shape,
which so bends and distorts the layers that they are
liable to be found in all directions; yet the lamellar
structure remains, however much broken, and, when ob-

MECHANICS. 253

tained in a suitable degree, contributes to give that orna-
mental surface called damask; and as all much elongated
iron, such as nails, wire, &c. shew a fibrous texture when
acted on by weak acids, this intermixture of oxide may
cause variation in that texture.

Now, the evils which arise from oxidizing the surface
increase with the size of the weldings; for the great length
of time the surfaces of large masses must be exposed to
an urgent fire to obtain the welding heat, will cause so
deep an oxidation that the union will be so much the
more difficult to effect ; for there needs a proportionate
increase of hammering to break through a thicker coat of
oxide, while the difficulty is increased of transmitting the
force of the hammer through larger masses to the welding
surfaces; so that the union is liable to be as much less
perfect as the mass is greater.

Accordingly, it is very common to meet with anvils
whose large beak, and sometimes other parts, are broken
off at their weldings; shewing how very imperfect the
union was, and that only at the edges, and seldom any
at the centre.

And as an acknowledgment, by a workman, of the
general deficiency of weldings, the Society, in 1820, re-
warded Mr. R. King for his method of avoiding them,
by forging the anvil-top, its beak, and opposite over-
hanging, in one piece, the base in another, and then
welding the one on the other ; so that in use this welding
is the least exposed to strain.

Again : We frequently meet with iron-wire that will
strip in two, like splitting a twig, and sheet iron that will
split or peel up in parts : all shewing either the inability
of producing complete union or of knowing when it is
effected.

254 MECHANICS.

I have thus far traced the unsoundness which accom-
panies welding : next I shall shew what takes place when
the weldings are quite sound ; and shall pursue it, if pos-
sible, to the obliteration of every sign of previous welding.

If we recur to the melted carburet of iron, which is
puddled or brought to a pasty state, and then hammered
at a great heat till it has lost all its carbon, and till every
part of the metal has united in one mass and become
malleable, and observe what takes place during the pro-
cess, we shall obtain a tolerably good hint for welding.
The iron is still considerably divided by an excess of
carbon, and probably by other impurities not easy to be
defined. These are fast oozing out, and the carbon is
burning at the surface, while the hammer is continuaJly
bringing portions of the iron into closer contact, so that
they unite ; and the pure metal being much stronger and
tougher than the crude, it will keep together and com-
bine with every particle of sound metal with which it
comes in contact while the crudities are exuding; and
this goes on, to the improvement of the quality, so long-
as any carbon remains ; for, during that time, perfect
welding must go on, because the carbon takes the oxygen
and leaves the surfaces pure or clean for union. Thus the
iron becomes one solid mass free from oxygen; for oxygen
and carbon mutually take each other away in the form
of gas, and leave the iron pure. Now, if ever so much
hammered after this, the iron may proceed to perfect
soundness and homogeneity ; but it will be at the expense
of the surface, which is detached in scales as it passes to
the state of oxide, in which state, as I have before ob-
served, it is incapable of uniting with metallic iron.

The conversion, however, of crude into malleable iron
is never perfect till the mass is so reduced in thickness

MECHANICS. 256

that the welding force of the hammer is felt all through ;
and then if all the oxygen is removed from within, it may
be obtained quite free from flaws (at least all but minute
ones) and is fit for use.

Having thus established some thickness, up to which,
with given tools, new iron may be obtained sound, we
begin to increase that size by welding ; and here begins
a new and very different cause for unsoundness, which
will require our farther notice.

Various means have been resorted to in order to pro-
tect the surfaces of iron from oxidation while in the fire,
such as sand, glass, salt, or any thing that will bear the
fire ; but these require to be scraped off* before the bars
are put together for union. Now, any portion of such
matter that is not removed is itself a cause of unsound-
ness, and when perfectly removed, a thin coat of oxide
forms on the clean surfaces and is shut in.

In order to obviate this latter evil, the iron should be
coated with carbon, as though it was to be case-hardened,
by immersing or laying it on a bed of carbonaceous matter,
kept very hot till it has imbibed enough to engage all the
oxygen that would otherwise attack it during welding :
moreover, this carbon blazing out at the surfaces, helps
to keep up the heat till they completely meet together.
The surfaces ought also to be slightly convex, that weld-
ing may begin at the middle (that part being liable to
receive less impression from the hammer), and proceed
gradually to the edges : thus the blisters that are liable
to be produced from the union of oxygen and carbon are
more likely to be beaten out. And more particularly
as the masses singly were as thick as the hammer can
govern, and now being doubled, its effect is lessened at
the welding surfaces, therefore more care is required in

266 MECHANICS.

the mode ofbringing them together; for unless each partis
kneaded together enough to adhere before the heat lowers,
some parts will only be beaten close without union ; and
all after-heatings affect the outside more than the joint.
Also^ if any air is shut in^ this likewise retards the union ;
but a convex surface gives the best opportunity for it to
escape.

Next, supposing the iron protected as usual, if, at the
moment the substance used for this purpose is well
scraped off, the surface is sprinkled with good cast-iron
or, still better, with cast-steel filings, this addition will
introduce carbon enough to engage the oxygen, and a
sound welding may be expected. Here the operation
is similar to the original reduction of the iron to the
malleable state ; for perfectly clean iron unites very readily
if brought into close contact at a suitable heat.

This process for uniting iron which otherwise was too
thin for welding has lately become public, and by some
has been called soldering, under the notion that the cast
iron melted between the surfaces.

Iron is too thin for welding if it cannot retain the
heat long enough under the hammer (or when the re-
maining good metal is not enough to squeeze through, or
mix up with the oxide) to unite, without being smashed
to pieces by the hammer ; or if so much flies off in scales
that the remainder becomes useless. But by the intro-
duction of highly carburetted iron, the oxygen is consumed
and the heat thereby retained a little longer; and the
metal becoming all good, or restored, there is rather an
addition than a diminution of weldable substance ; and a
much less use of the hammer will bring the whole into
union.

Now, seeing by this means much thinner iron is united

MECHANICS. 257

than formerly, it is desirable to apply the same process
as completely to large work. For, though the mechanical
obstacles appear so opposite, the opposing agent (oxygen)
is the same, and is nearly in proportion to the masses.

Small work, being quickly brought to the heat, is less
burnt than large masses ; for these, requiring much time
to arrive at the necessary heat, receive a deeper coat of
oxide. Small work, however, is more disfigured in pro-
portion, by the beating requisite to cause union ; whilst
in large work the quantity of metal between the hammer
and joint resists the operation.

In one case the work is worth nothing unless the iron
is preserved; in the other the joint is not sound unless
oxygen is? excluded, because it can never get so much
brisk kneading together as to overrule the enclosed layer
of oxide, and force union.

A third method I would propose for very large work,
is to protect the surfaces in the most convenient manner
while coming to the heat ; then carefully cleaning them
and laying a thin sheet of clean hot steel, that has been
highly carbonized, between them : this would secure a
very sound and equable layer of metal all through the
joint, the carbon of which, if more than enough to reduce
the oxide, would still not impair the soundness of the
metal, and consequently do no harm to the joint.

If by some of these means, either of excluding oxygen
or removing it in the process, we are enabled to effect
a sound union, we may add layer upon layer of sound
metal, till we have obtained the largest required mass.
Yet this may, perhaps, be only apparently sound; for
here another evil is liable to be introduced, which before
was stated to need our attention ; and if this is allowed
to enter, it will increase with every additional layer.

VOL. XLVIII. s

258 MECHANICS.

In order to explain this, let the layers be supposed
to be added in succession, all on one side of the first,
bearing in mind that we are using the thickest layers
which our hammer can govern.

When two are welded together, the upper one is
stretched most by the hammer ; it is therefore pulling the
under one with it, which, if it cannot do, it will divide
the exertion by curving a little to it. Now here at once
the mass is in an unnatural state of tension, the under
piece being stretched like a cord, while by its force it is
striving to bend the upper one hke a bow. Now, a third
layer added on this, will, in like manner, be stretched
more than the under ones; and while endeavouring to
stretch the second one like a cord, it becomes the
antagonist of the first, and would overcome its power to
bend the second ; but as its relation to the second is
precisely what the second is to the first, a strong tendency
to a curve is still the result. Now, on proceeding to
add the fourth, fifth, sixth, &c. layers, we are continually
increasing the stretching force on one side. Therefore,
these additional layers must either all bend, or the under
one be so stretched as to break or become unsound ; for
the farther the bottom layers are removed from the ham-
mer, the more do they act merely like an anvil, suffering
less and less alteration from the hammer.

Now, if instead of adding only on one side, we added
as many layers on the three other sides (leaving the first
as a central nucleus) while hammering the outside layers,
they would all be enlarged ; the inner ones serving little
more than the bflSce of an anvil, are, therefore, stretched
till they break.

This kind of mischief begins with the first welding,
and increases with every layer, and is greatly increased

MECHANICS. 259

by the quantity of hammering required completely to
form and finish the work. For the more the outside is
extended by the hammer, the more the inner part is
stretched ; and as this will bear but little of extension
without parting, the mass must become unsound. Agree-
ably with this, it has long ago been observed, that large
masses hammered beyond a certain degree lessen in spe-
cific gravity instead of increasing, and therefore it was
supposed that a large ball might be hammered till it
became hollow. Now, whether the layers be added in
succession, or are all heated for welding together at once,
the effect will scarcely be altered, for it will require as
much hammering fully to effect all their unions, and
there will be the same gradation of the hammer's effect ;
for the outside will be drawn out in length more than the
inner. This is shewn by trying to lengthen a cylinder
by a hammer rather too small; for then the centre not
lengthening equally with the part directly acted on by
the hammer, the ends will become hollow, and the force
of the extended outside will stretch the middle parts to
unsoundness. The use of enormously large hammers
will carry us farther in size before central unsoundness
takes place, but then the surface is the more furiously
smashed about.

It remains to offer means of preventing or diminishing
this central unsoundness ; for though in large work the
tools would be expensive, yet where the repetitions are
frequent, that is of less consequence ; and if one ship
was saved by the greater strength of its anchor, it would
surely be an outset against the most expensive tools.

The bars, which are to be united into one great shaft
or stem, should be all of one length and of the full width ;
then an anvil or bed of iron should be provided, so large

260 MECHANICS.

as to contain a recess of the same dimensions as the
intended shaft. Through the bottom at each end, and
flush with it, iron posts might stand so that they could
be raised by levers at their bottom when the work required
to be lifted out. Now, prepare two of the layers for
welding, and put them into the recess, and have a stamper
or welding hammer square-ended, as wide as the recess,
and let it drop from a carriage travelling parallel with the
recess ; in this case there is no room for extension
either in length or width, the metal can only close, and
as there is no lateral spreading to make way for the
hammer, its force will be directed more effectually down-
wards; but as its face would be too broad for its best
effect, two men, with a frame like a hand-barrow, might
carry a short cylindrical hammer-like punch under the
stamper, and distribute the blows equally all over. Where
this cannot be done, placing the hammer over an anvil,
with two heavy cheeks adjustable to the required width,
would prevent lateral spreading and lessen the evil,
though it could not prevent the longitudinal stretching.
But by whatever means the workmen can prevent all
extension under the hammer, soundness will be preserved.
The above process being performed on two layers, then
bring the mass thus produced to a welding heat, and add
a third layer, and so on in succession.

If a ball of clay or putty is rolled into a cylinder, it
will shew hollow ends, proving the outside to have
extended most; and the centre, being cracked by the
stretching, at last breaks the cylinder all through, and it
falls to pieces with cracks sharper than the softness would
lead us to expect. I have seen brass wire full of cracks
at which it would break, occasioned either by very hard
drawing or brittleness of the metal, and these breaks were

MECHANICS. 261

convex, fitting into concaves, shewing that the centre
had travelled faster through the hole than the outside;
for the convex was towards the hole before it had passed,
and from it when passed, and the last end of wire is
frequently hollow.

Now, seeing large masses are so liable to acquire
internal unsoundness, and the weldings also difficult to
obtain sound, I would, in constructing anchors, totally
avoid all weldings across the pull to Vv'hich they are
subjected when in use.

I would therefore take new iron that has only been
reduced to the malleable state, in pieces long enough to
forge into the complete form of the intended anchor,
without flukes, and thin enough to be quite manageable
under the hammer. Having provided a quantity of these,
I would weld together (as previously mentioned) by
carburet of iron sprinkled between them, a suflicient
number to give the required thickness, with some con-
trivance to prevent all extension, but should prefer a
recess capable of receiving the whole anchor. This would
produce square limbs ; and I should prefer chiselling off
the corners to hammering them in; for if they were
hammered in, it would put some parts in a state of
greater tension than others, and destroy that equal ease
throughout, which is so very essential in an anchor,
where every part should contribute equally to bear the
strain ; for those parts that are left in a state of tension
chiefly bear the pull, and must give way before the
others can come to their full tension ; therefore the
anchor breaks. Perfect soundness of the iron, an arrange-
ment of the weldings so as to be in a line with the
pull and ease of all the parts, or an equality of their
state and freedom from tension, are the three most

262 MECHANICS.

essential requisites in an anchor^ after its form has been
determined.

From these considerations, I think it will appear, that
many anchors are rendered unsound by the hammering,
and that the metal is liable to be put into such an un-
equal state of tension, that one part at a time only bears
the strain ; so that with the weight of a large anchor we
have only the strength of a small one ; and when this is
the case, can we wonder at their breaking ?

There remains yet another view in which the use of
large implements should be considered ; namely, at what
thickness can any given material be used with the best
advantage. This leads to the inquiry, whether there are
any limits in this respect as to thickness and length ; and
if so, what is their nature, and do any implements exceed
those limits.

Now, an anchor is the largest hook in use, and is
exposed to rougher or more violent usage than any other,
and that even in proportion to its greater size, on which
account it ought to possess every good quality of metal.
But I fear this inquiry will shew that the largest anchors
do exceed the limits of some of the most important
quaHties.

To seek for these limits, we must look at matter much
reduced in size, and compare it with the largest masses,
and we shall find a greater contrast than in size.

Small pieces fall from any height without breaking,
the blow they give or receive not being equal to the force
of cohesion between their particles. Now, on increasing
their size, the sectional surface to be broken increases as
the square, while the weight increases as the cube, there-
fore large bodies break from falling. .

If we sufficiently reduce the thickness of very brittle

MECHANICS. 263

materials, such as glass, they are found to be flexible :
glass may be drawn out so thin or fine as to bear twisting
together and forming thread, and it bears throwing about;
while the same substance in rods two or three inches
thick, will break rather than bear bending enough to be
discovered by the nicest measurement.

If we could have wire so small as to consist of one line
of particles, it would lose all character of stiffness, and
bear unlimited bending (such an imaginary wire may be
imitated by a line of small magnets between two large
ones) ; and if it consisted of three or four lines of particles,
no bending could remove them out of the sphere of their
mutual attraction; such a wire, therefore, would never
break. Now, the smallest soft iron wire bears so much
bending about, that it appears, in this respect, more like
lead than iron of large dimensions.

This flexibility continues so far, that soft iron wire,
about one eighth of an inch thick, may be bent double,
while cold, without fracture. Here Fig.i. Fig, 2.

the outer portion a b, fig. 1 , about
two thicknesses long, is bent and
stretched into the curve a b, fig. 2,
about double the former length ; the
fulcrum round which it begins to
bend is probably one-third within,
but it must ultimately remove to the outside, as will be
seen by the figure.

We may observe that a considerable angle is made
before there is much stretching : this is of the greatest
consequence, as it immediately engages a longer portion
of the metal to stretch and contribute to the bending ;
for where there is any curvature the metal is stretched :
this will be understood by bending wire into a spiral

264 MECHANICS.,

round a very small axis, when it will have a constant
succession of bending parts, and the whole outside will
be stretched. Now, after the angle is made, the outer side,
while bending and stretching, is forcibly pulled closer to-
wards the fulcrum, therefore the change of place of the
particles is not so great as it appears ; and the reason the
wire seems not to have lost any thickness at the bend, is
owing to the bulging out of the inner side, nearly making
up for what is lost on the other : the inner portion is so
hardened by the pinch, that there is a little lateral sliding
round it while making the bend.

Now, supposing we have begun with the thickest wire
that [will bend double, let us take some of twice that
thickness, the sections of the bending portion will contain
four times the surface, but the quantity of metal engaged
in bending is eight times as much ; therefore the amount
of its cohesion is now doubled ; and in consequence this
wire cannot be bent so much as the other without
breaking.

Next take a rod of ten times the diameter of the wire;
its sectional surface, the cohesion of which resists breaking,
will have increased a hundred times, while the quantity of
metal to be bent is increased a thousand times ; this is ten
times as much in proportion to the cohesion, and the lateral
change of place among the particles requires also to be
ten times as much : but we supposed the first wire to be
the thickest that would bend double, consequently there
was all the change of place among the particles that the
metal could bear ; but in this latter case, ten times the
former amount of change of place being required, with ten
times the proportional quantity of metal to the cohesion,
the rod must rather break than bend.

Steel or iron wire, 1-twentieth of an inch thick and a

MECHANICS. 265

foot long, will bear throwing about any how, and suffer
little from bending; but if, with the same proportions, the
thickness be one foot, the length will be 240 feet, and its
weight somewhere about 80,000 lbs. : this, it is evident,
could not bear falling from a height proportionally great,
or even being lifted from one end, like the smaller.

Thus the power of bending decreases as the thickness
increases ; therefore, in very large masses it is almost
nothing ; and this same reasoning applies equally to the
springing power of large masses.

Now, seeing that brittle bodies become flexible by
reducing their thickness extremely, so the most flexible
bodies become, as it were, brittle by greatly increasing their
thickness : therefore, I think we may come to this con-
clusion, that there is a given thickness at which metal
loses all power of springing or bending, and therefore a
blow powerful enough to require its yielding will be sure
to break it, because it cannot yield in any other way.

We may notice that, after a certain quantity of bend-
ing, the iron becomes so unequally stretched, that some
parts are rendered weaker than the rest; these cannot bear
the same exertion, and so must give way.

When a bar receives a blow, or takes one by falling
from a great height, it must either spring, bend, or break,
at the part struck, because the section there is the shortest,
as at fl?, fig. 3. Sup-
posing A the fulcrum
or blow, the sections
b c and d rapidly in-
crease in length, and
therefore soon limit
the portion whose strength has to resist or oppose the mo-
mentum or weight of the two ends, and when it bends.

266

MECHANICS.

the section a is most stretched, b c and d succcessively
less as they are longer; therefore a is the only part
stretched to the utmost; but when a curve is made by
a little bending, then there is just so much added that
can stretch like a, or at least like b.

It may, perhaps, be allowed here to give another view
of what takes place while bending. We will suppose a
bundle of square rods A A fig. 4 perfectly fixed together
at their ends, so that they cannot slide, to be forcibly bent
as to B B. In this case the rod marked o, or one near that

^^M

1^

place, would simply bend ; No. 1 below it, would thicken
and shorten; while the bottom one 2 would bulge out
making three bends like the letter M ; then Nos. 1, 2, 3,
and 4, above, would all have to stretch as well as bend,
and while their middle was stretching there would be a
little sliding round the fulcru^m or the two shoulders formed
in the lower bars, the upper bar stretching most, and the
middle part c in all four of them stretching more than the
adjoining parts, as is shewn by the sections in fig. 3.
Now, if instead of the rods being so small as to bear
bending, we suppose each rod to be the thickest that will

MECHANICS. 267

bear to be bent so much, it is evident they will not bear
the additional quantity of stretching and sliding requisite
to allow of their taking the form b b, for the parts a a
would have to descend iobb\ a portion d d, as wide as the
fulcrum is thick, is pulled straight just before rending,
and in all cases of bending that straightness is a signal
to go no further, for the bar will break.

Now, seeing the power of bending or springing de-
creases as the thickness increases, and that the weight (the
momentum of which endeavours to break any body which
falls suddenly on a hard resistance) increases as the cube
while the strength increases only as the square, I think it
will follow that anchors ought to increase more in thick^
ness than in size, in order to make up for their loss of
flexibility, and enable them to fall without breaking.

This leads us to inquire whether anchors are formed
so as to answer equally well their three principal services :
the first, to be sure of taking hold ; the second, to bear
the great strain and jerking to which they are subjected
while in use; and the third, to bear falHng; and sometimes
a fourth, the capability of being withdrawn from a good
hold.

With respect to the first, the flukes are probably as
broad as the strength of the anchor will bear ; for if they
were broader, they sometimes would be liable to be twisted
off when the ship swung round with the tide; and the
form of the hook is evidently good for taking hold, and
running that hold up to the shank; but as the fluke is
the chief resistance, the arm should begin to curve from
the fluke gradually into the shank : but the wooden stock
which obliges the anchor to take hold, offers as much
surface to the ground as the fluke does; this added to the
rising angle of the pull, prevents the whole anchor from

268

MECHANICS.

diving through soft ground to a harder bottom in which
it can hold. The power of diving might be given by
using iron stocks, made to turn in the anchor, with the
arras fig. 6 flattened, and round disks at their ends
turned inwards enough to make the flattened arms present
their edge to the sand or mud in such an angle as to be
able to dive. Again, it is not well to drop an anchor with
any additional weight, as that would tend greatly to break
it ; but if a stop were put on all cables three or four
fathoms from the anchor, and a weight like a collar
allowed to run down the cable to this stop whenever the
anchor was found to drag, it would cause such an angle
in the cable as to make the portion between it and the
anchor lie on the ground and pull horizontally : this
weight would also act like a spring, for when the ship
tugged at the cable, endeavouring to straighten it, it would
have to raise the weight before the anchor could feel the
tug, and then it would never be so sudden or so much
upwards as without a weight.

Fig, 5.

With respect to the second condition, the action of a
hook is the severest strain that metal can be put to ;
therefore, nothing but the most careful gradation of form,
regularly increasing the strength from the stock to the

MECHANICS. 269

bow and from the fluke to the shank, and curving the
shank better into the bow or arms with really sound
'metal, can distribute the strain equally.

Thirdly, the bearing a fall : an anchor is so formed
that it cannot drop on a hard bottom without strain some-
where ; and as we can only meet this by increasing the
thickness more than the size, and as with this thickening
it will be gradually losing the form of an anchor and
approaching that of a block, it becomes a query whether,
for the largest anchors, Parks's mooring block (which was
rewarded by the Society in 1818, Vol. 36) is not the best,
weight for weight, because it will dive till it finds hold,
and a stopper may be put on the cable to prevent it from
diving too deep to be easily withdrawn.

There are two somewhat similar operations by which
metal is fitted for our use that deserve our consider-
ation, because in one it evidently loses strength laterally,
although it gains it longitudinally. These are wire-
drawing and rolling out into plates: in one case it is
drawn through holes successively reduced in size ; in the
other it is drawn between hard rollers successively placed
nearer to each other.

This produces a character rather extraordinary for
metal, for it becomes fibrous; and this texture may be
satisfactorily shewn, by submitting a piece of it to the
slow action of weak acid. Some have attributed it to
crystals which kept their separate character while being
elongated. But if we carefully trace the progress of the
metal through the holes or between the rollers, we may
probably find sufficient cause for an elongated or fibrous
texture.

The metal is caused to pass through the holes by

270 MECHANICS.

pulling and not pushing; therefore the quantity reduced
each time must be small compared with the thickness
of the wire, because it is the mere tension of the reduced
part which causes a complete change of place among the
particles, or the squeezing of larger rings or short cylin-
drical portions successively into a smaller size, and
elongating them in proportion. This elongation takes
place in the hole where the metal is stretched nearly to
the utmost, and it may be considered as analogous to
drawing a cylinder or core out of a tube ; and this core is
so much stretched as to receive into itself the whole of
the outer coat, which thus sinking in becomes part of it
and follows by its cohesion. Now, as these successive
rings are too large to go through the hole, they cannot
contract and dive into that below but by a longitudinal
sliding of the particles, some sliding aft to let the others
close and proceed through the hole. Now, as the elonga-
tion is unlimited while there is room to reduce, the
centre would be torn asunder, did not the outer metal
dive and fill it up; then, as cohesion, or the power
which resists parting, is the cause of this, the whole
portion at the hole may be considered almost as in a
liquid state while passing through it. During this motion
and change of place among the particles, if there is any
sort of polarity in them, this is the time they are called
into action, and can arrange themselves, and that arrange-
ment must be longitudinal ; but the longitudinal sliding
of the particles to allow some to proceed first through the
hole, and the continual wedging of portions of the outside
into the central parts to supply the current of particles
through the hole, (for it is like a current) the particles
re-arranging themselves and following by mere cohesion.

MECHANICS. 271

appears sufficient to produce a tendency to longitudinal
fissures, by which the coherence is somewhat weakened,
although no real division takes place.*

In the flatting-mills the revolution of the cylinders
helps to carry the metal through ; in that it differs from
wire-drawing ; but though there are no side stops, the
metal scarcely spreads any thing in width, proving that
there is no pressure of the particles together laterally, but
that the mass is elongated just as the thickness reduces;
here the rollers pinch tight the reduced part and forcibly
carry it through ; while the thicker part, like a wedge,
refuses to follow, so it is stretched somewhat like wire-
drawing and the surface is driven back in waves. Here
the whole action is lengthwise, and each line extends
without regard to its neighbouring lines, so that its lateral
adhesion is weakened ; and, as a thin plank will bend
lengthwise, but will split rather than bend crosswise,
so will hard rolled plate metal. Annealing and ham-
mering once or twice till hard again, by distributing the
extension in all directions, removes this parallel fibrous
character. This change of character, according as the
metal is worked, shews that none is so strong as that
which is beat in a recess ; and I think no fibres could

• When reducing some thick wire in the lathe, to make a smaU screw
with a large head, it proved so unsound at the centre as not to bear
reducing to the required size ; it did not break off, but tore aside ; and
on twisting it about with the pliers, the whole of the reduced portion
appeared fibrous, so that the wire seemed to have a pith within it. In
this instance the centre must have stretched more than the outside could
supply while drawing. It would be desirable to try which would give
the soundest centre, — the reducing wire through the fewest possible number
of holes, or greatly increasing their number and bringing the wire down
very gradually ; then drawing some very slowly, and some as quick as
possible.

272 MECHANICS.

be discovered in such metal ; for, if I may use the expres-
sion, it would be one fibre spreading every way; in fact,
it becomes quite dense and homogeneal.

I have now endeavoured to shew that oxygen entering
with each welded surface deteriorates the iron ; and have
proposed some and pointed to other means which have
been found capable of avoiding it.

Also, that the difficulty of transmitting the force of
the blows to the inner surfaces increases with the thick-
ness, and therefore limits the thickness at which sound
welding can be obtained.

And, when the hammer is used for lengthening a mass,
the outside is chiefly elongated, from the effect of the
hammer regularly decreasing as the centre is farther from
the surface; so that, at last, it receives no impression, but,
with the mass below it, serves only the office of an anvil
merely helping the hammer to extend the outside, in
which case the centre always becomes stretched, and
thereby weakened, and sometimes torn asunder ; and
as each side while lengthening under the hammer endea-
vours to bend the mass, it causes an alternation of the
stretching or springing, so as to facilitate the cracking
of the centre.

Thus, while small masses are rendered dense and
stronger by the hammer, large ones suffer loss of strength
if very much hammered.

On Steel.

Steel is made by combining carbon with iron : it is
done by immersing the iron in carbonaceous matter at
a white heat and excluded from air, till it has imbibed
enough all through. It is made to receive more or less.

MECHANICS. 273

according to its intended use. Pure iron is not altered nor
increased in hardness by sudden cooling from a red heat ;
but the small quantity of carbon which is combined with
it to form steel, greatly increases its strength and tough-
ness, leaving it both malleable and ductile, and gives it
that peculiarly valuable property of becoming extremely
hard, if suddenly cooled from a red heat by immersion in
water or by any other means. With this first dose of
carbon it is called 7nild steely because it possesses all the
valuable properties of iron, with increased strength.

A larger quantity of carbon increases the hardness of
the metal, and makes it more brittle when hardened by
sudden cooling, and also renders it fusible ; therefore it is
called cast steel, and being less malleable, it is more diffi-
cult to work.

When steel is made without fusion, by immersing it
in carbon, the process is called cementation ; and it is called
blistered steel, because, while the carbon is entering, it
meets with oxygen, or hydrogen, or other impurities,
which it causes to become gas or vapour, and this blis-
ters the steel. There is a considerable quantity of steel
brought to market with defects which appear to be the
remains either of these blisters or of the weldings that
follow.

Cast steel, being made by fusion, admits of an equal
distribution of carbon, and of the escape of every particle
of gas, vapour, or any other substance not compatible
with it at so great a heat, leaving only such as can com-
bine with it ; therefore it is the only steel we can be sure
of obtaining quite sound.

The question whether steel contains any thing besides
carbon and iron, is chemical. A good workman only re-
quires metal perfectly free from flaws, and quite homo-

VOL. XLVIII. T

274 MECHANIGS.

geneous, and that will harden at the lowest heat ; for this
last test supersedes all others in proving the goodness of
steel and its fitness for the best purposes.

The soundness of cast steel renders it most desirable
to use ; but the excess of carbon makes it too harsh, and
therefore more difficult to work ; but by frequent heatings
and hammerings it may be reduced to the state of mild
steel, the excess of carbon burning out while forging, and
then it must be the best steel for general use. Yet shear
steel predominates in the market, and more particularly
in the various manufactured tools, probably from being
cheaper. Therefore it is desirable to discover its faults,
the better to cure them.

Perfectly pure iron cemented in pure carbon would
probably make steel without blisters; but as in practice
blisters are produced, there must still be oxygen in the
iron, or some other substance that is turned into gas or
vapour by the addition of carbon. Now, the blisters are
most on the outside, but why not equally throughout?
The answer is, the strength of the metal resists the as-
sumption of the gaseous or vaporous state. Is not, then,
the blistered outside purer or better steel than the centre?
for though the carbon penetrates to the very centre, is it
not obhged to combine with the oxide of iron and other
impurities in the solid state, and remain so, the strength
of the metal forbidding the change by which it would
have a chance of escape ? Again, if permanent gas is
formed in any of the small bUsters within, how is it to
escape ? the hammer may greatly reduce the size of such
blisters ; but will there not be small flaws or cavities re-
maining? If, in addition to this, the iron contains the
bases of the earths or alkalies, or the earths themselves,
these latter may be reduced by the joint action of the

MECHANICS. 275

carbon and iron, and then being vaporised by the heat,
the metal may be blown info blisters. In this case the
cavities would probably be coated and thus rendered un-
weldable; these flaws would then spread with the metal
while working into plates or rods, and the mass, though
close, would not be united.

The blistered steel must be unequally carbonised, the
outside containing most; it is fitted for the market by
drawing out to greater lengths by the hammer, then fold-
ing double, and welding together again; then drawing
out, and again welding several times. This mixes the
various parts together, and thereby distributes the carbon
more equally, and condenses the metal ; it is then called
sltear steel, and considered fit for use. Now, these weld-
ings are liable to introduce flaws, — first, by imperfect
union ; secondly, by the carbon burning out of the sur-
faces that meet together, giving a stratum either of iron
or of steel with less carbon ; and these being softest, would
give way during the extension of the steel.

Now, whether from original blisters, or similar cavi-
ties from imperfect welding, such defects do very largely
accompany this steel ; and it is a question, whether any
process short of fusion can totally remove them. Yet
long-continued forging greatly improves the soundness
and homogeneity.

Besides flaws, there is another defect often met with in
steel, it is said to have pins ; for, when filing or turning, it
appears to have knots or pins in it harder than the rest of
the metal. A similar defect occurs in brass, when, through
the negligence of the founder, it contains iron or steel
filings ; also in wood, with hard knots : but this evil is
worst of all when it occurs in steel, which naturally is
hard to work. I have met with it in all degrees, from

276 MECHANICS.

mere harshness up to real hardness; so that while turn-
ing in the lathe, these parts would remain projecting out,
and damage the tool rather than be cut away. I have
filed some off, and then found their hardness nearly
approach that of the file. This inequality of hardness
causes the work and tool, though held by the rest, to
spring away from each other, rendering it difficult to
turn true. I think it is often caused by portions of the
metal being over-steeled, i, e. so completely charged with
carbon as not to soften by slow cooling. Another cause
is stated by Mr. Clement, who broke the steel across
these pins, having filed away the back to render it weak
enough to part at the right place, when he found a cut
or division, on which account he attributes the flaw and
its hardness to oxide of iron, which prevented the parts
from welding together ; and as oxide of iron is known to
be harder than steel, by its being used for polishing that
substance, it most likely will cause hardness where it
occurs. It would be desirable to ascertain the state of
the surface of the deepest blisters, to know whether they
are alloyed or oxidised, or in any way differing in their
state of carbonization from the most solid parts. When
such spots occur from oxygen, the adjoining parts must
be iron ; for there will be a gradation from oxide through
iron to the steel, therefore the circumference of such a
spot would be softest.

An excess of carbon renders steel harder and more
brittle, therefore inequality is liable to occur. This is
illustrated by iron hardened while casting ; and of that
which is intended to be soft, portions are frequently found
hardened by the wet sand : these parts near the outside
break with a more glassy fracture than hard steel. Good
steel, hardened by plunging from a red heat in cold water.

MECHANICS. 277

will always become soft by slow cooling from such a heat ;
but this hard cast iron does not : it requires burning several
hours at a red heat, and must not be smothered by fuel
to prevent decarbonizing, but, on the contrary, should be
exposed to the current of air, that some portion of the
carbon may burn out while the remainder has a tendency
to equalize itself; then, if slowly cooled, it is found soft.
Now, the knots or pins in steel are not softened by slow
cooling, and to burn them out would spoil the steel, it
having no carbon to spare but in the pins (supposing this
to be the cause of their hardness) ; and keeping it hot in
close vessels will not produce equality sufficient for any
good purpose. Even cast steel is liable to long veins of
harder portions than the rest. All these defects shew the
need of farther attention to the improvement of steel, and
for this purpose two sorts of hammering are requisite.

The first should be at a forging heat, to knead the
parts, and keep them moving among themselves almost
like a paste; and should be continued till the different
qualities are not only intimately mixed, but, if I may use
the expression, really dissolved in each other, producing
perfect homogeneity; for the carbon being thus spread,
will discover every particle of oxygen, and (it is supposed)
will expel it, and the metal will be rendered as sound as
we can expect from cemented steel ; for if free from
oxygen ancj all alloys, except the carbon, there is nothing
to prevent a perfect union of all the parts while under
the hammer. Good steel consists of that proportion of
carbon and iron which forms the strongest and toughest
compound; each best portion, therefore, when brought-
into contact by the hammer, remains so, and resists its
force the more from the greater cohesion of its particles ;

278 MECHANICS.

hence the redundant or deficient portions suffer most
kneading, till they are all equalized, and the cruder im-
purities are either beaten out, or formed also into a homo-
geneal compound with the whole mass.

Having thus far obtained sound steel, it is yet by no
means in a good state, being very unequal in density, and
in a state of distraction — some parts close and dense,
others squeezed out, and some nearly rent asunder,* there-
fore a second hammering is necessary at a particular heat»
and under circumstances as similar as possible to those
described in the former part of this paper ; such as re-
cesses in the anvil, or in blocks laid thereon, suited to
the shape of the steel.

For this purpose the metal is first brought as near as
eligible to the size in which it will be used, and then is
to be hammered in order to close and condense the metal
equally and to the utmost all through, and yet leave every
part in a state of rest and ease — a condition extremely
essential for good springs, sword blades, musical wire,
and every thing that has to vibrate, or act by its tension ;
for if one part is denser than another, or more at ease, the
weaker parts will have most, if not all of the play, and will
soon break. And that this is not so often noticed to be the
cause as it should be, is shewn by thecrudeness of the cure.
If a spring breaks, it is frequently replaced by what is
called a stronger, that is, a heavier one, losing a portion
of the play, and what remains being still upon the weakest
parts ; so that though it may last longer, yet it ultimately
breaks. This may also account for the breaking of axle-
trees that have appeared sound ; for they are subject to
such violent shocks as to require every part to yield
equtibly ; for the axiom, that the strength of the whole

MECHANICS. 279

is only that of the weakest part, rendered less so by the
weight and stiffness of the stronger parts, is as truly
applicable to springs as to ships.

This second hammering is also to prepare the steel for
receiving the utmost hardness, by that peculiar property
which it alone possesses — namely, that of receiving a
brittle hardness when suddenly cooled from a red heat ;
but though this hardened steel is brittle, its toughness
greatly exceeds that of any other brittle substance.

It is this quality that makes it completely the master
metal, the one by which we give shape and form to all the
others, and fit almost every thing for our use. I must
here observe, that this hardness cannot be given in part,
it must always be given in full ; and so true is this, that
in a piece of steel, part of which is hard and part soft,
no gradation of hardness can be detected ; the soft parts
adjacent to the hard ones being quite as soft as any
others ; indeed, so much so that they have been thought
by some to be softer than if slowly cooled. This appear-
ance may be accounted for in the following way : Suppose
a rod of well-hammered steel is heated at one end for
hardening, there will be a gradation of temperature from
the coldest to the hottest end, and the annealing, or re-
duction of that hardness which it has received from the
hammer, will be in proportion to the heat, consequently
the rod will be softer and softer towards that end where
the heat is applied. On plunging the bar into cold water,
that portion which had become hot enough to harden
becomes quite hard, and close adjacent to it will be
found that part which having been the most annealed,
will bear twisting and bending more than any other. But
though this hardness must always be given in full, it
can be let down in any assignable degree, that is, a

280 MECHANICS.

portion of its brittleness may be removed by a moderate
heat, and a greater portion by more heat, and so on, as
the purposes may require. This is called tempering;
and if hard steel be brought to a red heat and suffered
slowly to cool, it becomes as soft as if never hardened ;
it is then ready for re-working or re-hardening. This is
called softening, and is distinguished from annealing,
which is the same process of slow cooling but applied
to steel, iron, or brass, merely to remove all mechanical
condensation, whether by hammering, flatting, wire-draw-
ing, bending, or any other work; for if metal has been
altered in shape by the hammer or other work as much
as it will bear without breaking, then by annealing, it will
be softened, and may again be bent or altered as much
more ; and so on, as often as requisite. Now, as different
degrees of heat remove different degrees of condensation
received from the hammer, and a white heat removes all,
it is of great consequence to harden from the lowest pos-
sible heat, in order to retain as much condensation as may
be ; and it is a fortunate coincidence, that the greater the
condensation, the lower is the heat from which steel will
harden, and the stronger and tougher will it be. But
should this condensed metal be once over-heated, it will
no longer harden from that lower degree, but only from
a heat near that to which it has been raised ; its conden-
sation, with the advantages dependent on that quality,
can only be restored by re-hammering. The lowest heat
at which steel is generally brought to harden is a dull
red, just visible in daylight. Therefore, to be safe, a
dull red, only visible in the dark, is chosen for the ham-
mering. At this heat also it can be kept coated with car-
bonaceous matter, instead of having any burnt out, which
is of particular consequence. For this purpose a smith's

MECHANICS. 281

forge, in rather a darkish place, having its flat bed even
with and near the anvil, is kept smothered v^^ith small
fuel, the bellows being used only enough to keep the
fire alive, so that the gas or smoke cannot burst into
flame. The several pieces of steel are laid a little in the
half-kindled fuel, enveloped in smoke; this coats them
with carbonaceous matter, which the hammering heat can
hardly dispel. They are brought in succession from the
fire to the hammer, and back again to the fire when too
cool ; the hammer moves quick, and every part of the
steel in close succession is slowly passed under it, and
then (the position being changed a few degrees) it is again
passed under, and so on, till it has been hammered in
every direction, being often reheated ; and this is con-
tinued as long as the workman thinks it can be of ser-
vice. Experience teaches him to know by the feel, the
sound of the blows, and the lessening degree of impres-
sion, when the steel is hammered enough. By such
means, and an honest zeal for the goodness of his work,
Mr. Walby, whom the Society in the year 1804 rewarded
for his hammer, produced the best trowels that probably
were ever made ; their toughness, spring, and elasticity,
seemed carried to the utmost. At this heat the metal
spreads much less; but yet, unless confined within a
recess so as entirely to prevent spreading, the outer band
will be so bad as to require cutting off*, it will be so
stretched and rent by the pressure of the inner portion.

From what has now been stated, it will be seen that
there are two causes for the failure of steel, more particu-
larly if used quite hard. The first is an unequal com-
bination of carbon with the iron, and this is more or
less the case with all steel till sufficiently hammered ;
the toughness of some parts, the brittleness of others.

282 MECHANICS.

and the different states of tension hence arising, render
the metal very liable to crack, if not in the hardening,
yet afterwards, when forcibly struck as medal dies are.

The second cause is bad hammering ; for I think I
have shewn above, that however much hammered, it yet
may be left in a most violently conflicting state, some
parts girded and pressed, while others are as powerfully
strained, almost to breaking; and if hardened in this
state, how can it be expected to stand ? and thus springs,
though very equally formed, may be very unequal in
their strength.

Well-hammered steel requires the least tempering to
give it the necessary degree of toughness; but when
hardened from a great heat, it loses all the previous con-
densation from the hammer, and with it so much strength
that the toughness disappears, and the same hardness with
less strength shews itself in brittleness. Steel, therefore,
of this inferior strength, requires more letting down by
tempering, to arrive at a degree of toughness enough to
bear using. Such tools are too weak and soft for turn-
ing steel or iron, and do not stand long for any purpose.

I believe there is a given degree of cold to which the
steel must be brought in a given time to cause hardness,
also a given degree below which steel will not harden
from ; and that all farther increase of the heat only
weakens the steel, and all farther increase of coldness
serves only to harden a greater mass, or a deeper coat
of a large mass, by cooling it in the required time.

Steel much over-charged with carbon is too harsh or
brittle to receive all the improvement that hammering
would otherwise give it ; but by choosing the most
malleable steel, already sound, and hammering it at a heat
so low as to be capable of holding a coat of carbonaceous

MECHANICS. 283

matter, it will imbibe the carbon so slowly and in such
small quantities during each hammering, as to enable the
workman to bring it up to the fullest charge compatible
with sound hammering ; and so far carbon must improve
the strength ; but beyond that, brittleness comes on, and
it refuses to receive any condensation from the hammer,
by which alone toughness is given. It is of no use
seeking for hardness unaccompanied by toughness, for
we should have to let it down to prevent breaking ; yet
many believe the hardness to be increased by coming up
to this brittleness ; therefore, taps, dies, and turning tools,
in order to increase their hardness without losing tough-
ness, are allowed to receive a little more carbon on their
surface, by putting them in red hot carbonaceous matter
for the hardening, the toughness of the metal within
remaining the same. For this purpose animal charcoal
appears best. I have chiefly used burnt leather, which
seems capable of a sort effusion on the surface of the
steel while coming to a red heat, and therefore imparts
the carbon much quicker than wood charcoal ; and in
all cases of hardening it is best to heat the steel in close
vessels, or a case of some sort, to protect it from air
and prevent carbon burning out of the surface. Water
damages the surface at the moment of plunging ; to pre-
vent this, small articles are frequently plunged in oil or
tallow, which has the reverse effect, for it rather restores
the surface. Files and other tools that do not admit of
being sharpened are left quite hard, also tools for cut-
ting steel ; but in all other cases, to prevent the risk of
breaking, the extreme hardness is removed by tempering
more or less, as the intended work will allow.

Various methods have been resorted to, in order to
measure the exact temper, and more particularly for long

284 MECHANICS.

thin articles, such as watch-springs, which are diflScult
to heat uniformly. Melted metal, the fusing point of
which is just under the right temperature, has been
used ; but if the mercury or other melted metal be in
sufficient quantity, and the heat be measured by a ther-
mometer, it will secure accuracy : the article being moved
about in this till of the same heat in every part and all
through, will be well tempered, let the shape be what
it may.

Heating the articles in oil till the smoke rises copi-
ously gives a good temper for tools for brass-work, and
a still lower temper is given when the oil catches fire —
this is called blazing off; but for articles of any substance,
colour is the simplest and most direct criterion. The
hardened steel is ground clean along one side and kept
perfectly free from greasiness ; it is then heated, in pre-
ference at the side of a fire to avoid smoke, carefully
watching the bright part until it becomes of a straw-
colour, it may then be cooled in water to prevent the
spread of heat from parts not cared for ; this temper suits
tools for brass-work ; but if heated till it becomes brown
bordering on purple, it suits tools for pewter and very
hard wood ; after this it becomes blue, indicating a
suitable temper for the softest carpenter's tools, table-
knives, and springs ; it is just low enough to bear
filing, and in thin pieces will bend a little before it
breaks ; if heated beyond this it turns grey and is almost
visibly red in the dark. Very thin springs are observed
to be stiffer while the blue colour is on than when cleaned
off; and on reblueing them, they regain their stiffness
although there is no alteration of the temper ; such
springs are therefore preferred with the blue colour on.

These colours may be given to hard or soft steel, and

MECHANICS. 285

when cleaned off, the same heat will always restore them.
The steel and screws of watches are generally blued for
ornament; the other colours, when given merely as a.guide
in the tempering, are cleaned off. I have found the slow
conducting power of lamp black a useful agent to preserve
particular portions from being hardened with the rest.

It is sometimes desired to harden the neck of a man-
dril and not the screw, lest it should break when roughly
used : this may be done by an iron tube fitted a little
way on the neck of the mandril and ramming the space
between full of lamp black, so as completely to envelop
the screw, then shut it in by a disk to serve as a wadding;
the mandril being then made red hot and plunged in
water, the exposed part will be hard, and the covered
screw will remain soft.

Steel hardens as well under cover as if exposed, pro-
vided it can be cooled in the requisite time ; small articles
for watch-work have accordingly been hardened quite
clean, and their brightness preserved, by filling with them
a brass box capable of being shut air-tight while in the
fire, and then plunging the box with its contents in the
water. On opening the box when cold, the articles are
found hard and clean. Very fine drills or wire may be
hardened from the flame of a lamp or candle, by merely
shaking them quickly in the air, as that will cool them
soon enough. Large masses require rapid motion in
cold water to enforce their cooling in the requisite time.
The largest masses that can be hardened are best done
under a waterfall, the force of which beats away the
steam as fast as formed, and keeps the surface cool
while the central heat is escaping through it; and they
should remain in this situation till quite cold through-
out; if taken out sooner, the central heat will spread

286 MECHANICS.

to the inner side of the hard shell and expand it, while
the outside may be cold and therefore will be liable to
burst.

I met with an evident case of cracking, when hard-
ening some badly hammered steel which also was un-
equally carbonized. The pieces were ovals, one inch long,
filed out of steel bars one- tenth of an inch thick ; they
were then hammered, which condensed the middle and
stretched the outside ; in this state they were heated in a
crucible full of charcoal powder (which probably carbo-
nized the outer edge most) and hardened by plunging
in water ; this cracked them all in the middle, the most
condensed part, and none of the cracks extended to the
outside.

Stamps and medal dies are an important application
of steel ; and to enable them to be hardened sound, and
to stand in use when hardened, requires the metal to be
in a state of perfect ease, the result of equal condensation
all through, and this can best be secured by hammering
in a recess.

For this purpose it would be very desirable to ascertain
by experiments the greatest thickness that can be hard-
ened all through ; also the state of carbonization most
favourable to the greatest thickness ; or, what is the
precise difference in this respect between the most mild
steel and that which is highly charged with carbon.
Likewise, in very large masses, what is the greatest depth
from the surface that can be hardened, and v\/hether
greatly hammering it would cause any difference in the
thickness of the hard crust. This would of course require
the pieces so hardened to be afterwards broken to examine
the interior; and where a soft nucleus occurred it would
require to be ground flat, to shew what sort of boundary

MECHANICS. 287

there would be between the hard and soft parts; for it
would be rendered visible by the very different texture
which grinding produces in hard or soft steel, and there
must be some difference in the durability of a block,
according as the boundary is well or ill defined.

We are so familiar with small flaws in articles of iron
as scarcely to notice them, for their cheapness enables us
to use the good and throw away the bad, — such as nails,
brads, screws, &c. ; but our attention is drawn to large
flaws, from the serious consequences likely to attend their
giving way.

But in steel nearly the reverse takes place, as it is
used so much smaller than iron in things of consequence ;
therefore the smallest flaws are frequently of as much
importance as the largest, for they occasion the breaking
of tools, frequently spoiling the work; and when small
work is nearly finished, hidden flaws destroy it; they
also help to crack large masses when hardening.

But when the utmost efforts of the human mind are
transmitted to steel plates, from which the delights of
peace and civilization are spread abroad, it becomes of
the greatest consequence to avoid every flaw, and even
chemical dissimilarity; for though the metal be sound,
it may be so unequal in its nature as to etch very badly,
and the smallest flaws will spread very broad while
drawing into thin plates; and when the etching or
graving reaches through to such a flaw, the work is
spoiled, laminae of engraved work frequently coming off,
as engravers have already seriously experienced.

For such works, therefore, cast steel should be em-
ployed : it should receive an extra degree of forging, to
equalize most perfectly its composition and reduce it
to a mild state, for on this depends good etching; then

288 MECHANICS.

the surface should be watched and kept clear while
reducing it to a useable thickness, that no flaws may be
beat in: all welding must likewise be avoided, for fear
of shutting in flaws; and if rolled into plates, it will
require good hammering afterwards to restore the strength
or soundness which it loses laterally while rolling. The
especial reason why the primary forging should be long
and carefully performed, is, that the etching is liable to
be rough or smooth, as perfect homogeneity is more or
less obtained, and it will be cloudy if different parts of
the plate differ in their dose of carbon.

It would be desirable for engravers to make them-
selves well acquainted with the difference of action of
the same acid on pure decarbonized steel, on ordinary
steel, and on highly carbonized steel, and also on steel
in the softest as well as in the most condensed state ;
for they would thus judge better whether the defects
were in the metal or the acid.

Cornelius Varley.
1, Charles Street, Clarendon Square,
Somers Town.

No. VII.
MACHINE FOR WEIGHING COALS IN SACKS.

The Large Silver Medal was presented to Mr. James
Braby, of Duke Street, Stamford Street, for his
Machine for weighing Coals in Sacks.

A BILL having been introduced into the late ParHaraent,
for the sale of coals in the metropolis by weight, instead